Performance of a DIY DSLR cooler

[michaelmorris]

I am now nearing completion of my project to build a DSLR cooler based on the cooling mechanism and electronics from a small beer fridge.

I decided to test the efficacy of the system in cooling the interior of a camera whilst taking a typical sequence of deep sky images. 

The cooler contains a Canon 1100D DSLR camera.  The Canon 1100D records the temperature of the interior of the camera in the EXIF data for each shot.  APT camera control software has the option to record this temperature as part of the file name for each image taken.  The camera cooler has a small digital thermometer probe inside the box, a record was kept of readings from this thermometer.

I used APT software to set up a sequence of dark frame exposures of 200 seconds with 20 seconds pause between each exposure. 

The results of test are shown in the graph below.

As expected, the EXIF temperature initial rose as the heat generated by the operation of the electronics in the camera warmed the camera interior from an initial temperature of 26^oC to around 30^oC over the first 11 minutes of operation.  This rise may have been exacerbated by the heavy insulation around the cooler.

The graph shows that the cooling circuit initially provided rapid cooling of the air within the case of the cooler, but this took some time to have any significant effect upon the interior of the camera. 

The EXIF temperature record shows that it took 20 minutes for the cooling of the case to start having a measureable effect upon the internal temperature of the camera.

Ambient temperature stayed remarkably steady throughout the length of the test, with a starting temperature of 26.2^oC, falling just 1.6^oC to 24.6^oC by the end of the test.

It took 80 minutes of cooling to the DSLR to stabilise at around 22/23^oC, about 2^oC below ambient and 7/8^oC lower than the peak EXIF temperature.   Taking into account the 1.6^oC fall in ambient temperature over the test, it is concluded that the cooling system resulted in a fall in internal camera temperature of between 5.5^oC and 6.5^oC.   After the initial rise in camera temperature, the difference in temperature between the air temperature inside the case and the temperature inside the camera stayed relatively stable at between 8.5^oC and 11.5^oC.

After 130 minutes of cooling, the DSLR was turned off for 40 minutes.  The temperature inside the casing fell rapidly, falling by 2.1^oC over the 40 minutes of the test.

The camera was then turned back on and the imaging sequence recommenced.  The initial image of this second sequence of imaging showed that the temperature inside the camera had fallen to within 1.6^oC of the temperature in the casing.  However, over the next 27 minutes the temperature of the air in the camera steadily rose back to 20^oC, and the 7.7^oC difference in temperature between the casing and the inside of the camera was restored.  This supports the supposition that the DSLR was contributing around 9-11^oC to the temperature inside the camera.

Effects of temperature upon image quality

Dark frames were converted to 16 Bit TIFF format and each subjected to identical uses of the ‘Curves’ function in Photoshop CS5.

The effect of cooling on thermal noise was quite noticeable, with much lower levels of noise in dark frames taken after 90 minutes of cooling.  For comparison, the level of noise apparent in the frame taken immediately after the after turning the camera back on after 40 minutes of cooling with the camera turned off showed an even greater reduction in noise levels.

Conclusions

The DSLR cooler seems to provide limited cooling of the interior of a Canon 1100D camera.  After 50 minutes of cooling the camera interior had fallen 4^oC, with the cooling effect extended to 6.5^oC after 100 minutes.  This provided a small but measurable beneficial effect upon the levels of thermal noise apparent in dark frames.

It would be interesting to repeat this experiment in much cooler conditions where the beneficial effects of active cooling could be significantly less. 

[Knighty2112]

Well done. Looks like a result!  

[Adam J]

Nice, do you have any pictures of your design?

[michaelmorris]

2 hours ago, Adam J said:

Nice, do you have any pictures of your design?

No, not yet. I hope to finish the whole project off before the way end of next week, so I might get a chance to test it before I go to Kelling Heath star party at the end of the month.

 

[Stub Mandrel]

Looks a useful decrease - with night-time temperatures of 10 degrees and below your cooking on gas!

[michaelmorris]

Personally, I'm not convinced that I'll get the same degree of cooling when the ambient temperature is considerably lower than the 24 - 26oC I had during the test.  The problem is that the electronics on the camera seems to add about 10oC to the temperature inside the heavily-insulated cooling box.  If the ambient temperature fell to somewhere near zero, it might be more beneficial to leave the camera body outside the cooling box.

[michaelmorris]

On 08/09/2016 at 20:42, Adam J said:

Nice, do you have any pictures of your design?

Sorry, I misread your posting.  Pictures of the design can be found at

 

[Starflyer]

On 13/09/2016 at 23:43, michaelmorris said:

Personally, I'm not convinced that I'll get the same degree of cooling when the ambient temperature is considerably lower than the 24 - 26oC I had during the test.  The problem is that the electronics on the camera seems to add about 10oC to the temperature inside the heavily-insulated cooling box.  If the ambient temperature fell to somewhere near zero, it might be more beneficial to leave the camera body outside the cooling box.

I used to get 10 - 12C below ambient inside my cooler box.  The EXIF temperature would stabilise at around 15C above the temperature inside the cooler box, this was with 600s subs with a 30s pause in between.

So at 5C ambient (autumn / spring temps), I'd get -5C inside the box and an EXIF reported temperature of 8C to 10C.  EXIF would start close to the temp inside the box and rise over the first 30 to 60 minutes.

My 450D showed the typical horizontal banding noise problem at EXIF temps above 18C, so the cooling reduced the overall noise effectively and controlled the banding problem for most of the year. I sold also built up a library of temp matched darks using EXIF temps and these were extremely effective.

 

Cheers,

Ian

[Stub Mandrel]

How do you read the EXIF temperatures for a 450D?

I have an exif viewer that works on JPEGS but it refuses to recognise the RAWs!

[Starflyer]

I used to use Kuso EXIF Viewer;

http://www.softpedia.com/get/Multimedia/Graphic/Digital-Photo-Tools/KUSO-Exif-Viewer.shtml

[Gina]

I use FITS Liberator from NASA to view the EXIF data in my FITS files from astro camera.  Don't know if it works with DSLR files though...

[Stub Mandrel]

Thanks both.

[Stub Mandrel]

Interesting, although a long run of 100 darks varied between 20 and 26 degrees they were nearly all at 22 degrees, as are my M33 lights.

Ten degrees of cooling should provide a useful improvement.

[Jokehoba]

23 hours ago, Stub Mandrel said:

How do you read the EXIF temperatures for a 450D?

I have an exif viewer that works on JPEGS but it refuses to recognise the RAWs!

I use Picmeta's Picture Information Extractor (PIE free). Good for other things too.

http://www.picmeta.com/download.htm